|
 |
The origin of language, part deux | ||
Tags: Linguistics | Press Article | ||
Share with:
| ||
 Language Learning Forum : General discussion | ( Topic Closed) |  |
| Juаn Senior Member Colombia Joined 5143 days ago 727 posts - 1830 votes  Speaks: Spanish* | Message 1 of 10 17 April 2011 at 5:59pm | IP Logged | |||||||||||
|
Since the other thread was closed I thought I'd open a new one, given how this topic I believe is of interest to us, to be discussed this time without reference to subjects "forbidden" on this forum.
Here is an overview of the two recent articles that sparked the controversy: The evolution of language: Babel or babble? 1 person has voted this message useful | ||||||||||||
| Merv Bilingual Diglot Senior Member United States Joined 5071 days ago 414 posts - 749 votes Speaks: English*, Serbo-Croatian* Studies: Spanish, French | Message 2 of 10 17 April 2011 at 7:31pm | IP Logged | |||||||||||
I've gone through both the articles and the supplementary data. We all have our biases. Atkinson does show some correlation, but the Pearson correlation coefficients are very weak, in the range of as low as r=~0.15 all the way to r=~0.45 or so. In the sciences I work with, those values would be unacceptable for publishing. Another criticism of Atkinson's work is that he chooses to focus only on phonemic inventory, but I wonder if that was based on an unbiased assessment of all possible measures of language or simply the only one that supported the out-of-Africa narrative. In other words, if he assessed grammatical complexity, would he get the same result? Many languages with notably complex grammars, including those of the Indo-European, Uralic, North Caucasian, and Native American language families - are far spread out of Africa. To counter the argument that this is a result of innovation, most will agree that at least in Indo-European, it is the older and more archaic languages that tend to be more complex, i.e. Sanskrit, ancient Greek, Latin, etc. as compared to their modern descendants. The tendency is towards loss of complexity, not the reverse. Of course, comparing grammars is much harder than comparing phonemes, but I am sure there are ways to do it, as for example breaking down sentence formation into algorithms and seeing how many steps are required to get to the proper sentence. I will not comment on religion, but I do hope this forum is not so hostile to the religious that it will reject my scientific criticisms of one of these studies. 3 persons have voted this message useful | ||||||||||||
| tractor Tetraglot Senior Member Norway Joined 5251 days ago 1349 posts - 2292 votes Speaks: Norwegian*, English, Spanish, Catalan Studies: French, German, Latin | Message 3 of 10 17 April 2011 at 8:37pm | IP Logged | |||||||||||
Is this true at all? 3 persons have voted this message useful | ||||||||||||
| Merv Bilingual Diglot Senior Member United States Joined 5071 days ago 414 posts - 749 votes Speaks: English*, Serbo-Croatian* Studies: Spanish, French | Message 4 of 10 17 April 2011 at 9:43pm | IP Logged | |||||||||||
Are you quoting from the paper? My suspicion, although with little evidence, is that it is not. To give you an example: the Caucasus. Based on what I can find on wikipedia (my own counting and what the articles actually say): Armenian (Armenian branch of IE) has 36 phonemes Georgian (Kartvelian) has 33 Chechen (Northeast Caucasian/Caspian) has 84-104 Ubykh (Northwest Caucasian/Pontic) has 86 Turkish (Oghuz branch of Altaic) has 32 Azeri (Oghuz branch of Altaic) has 34 Talysh (Northwestern Iranian branch of IE) has 33 Ossetian (Northeastern Iranian branch of IE) has 40 Farsi (Southwestern Iranian branch of IE) has 29 Russian (Eastern Slavic branch of IE) has 45 (from what I could tell, counting palatalized consonants as separate from nonpalatalized) Kurdish (Northwestern Iranian branch of IE) has 38 Lezgin (Caspian) has 60 Abkhaz (Pontic) has 60 You are free to look up the population sizes that speak each of these languages natively and plot phoneme # vs. population, and I bet you wouldn't see much of a correlation. Most of these languages float around 30-40 phonemes, with the major outliers (the NE and NW Caucasian languages) have many more, but also being comprised of very small populations, often living historically in isolated mountain villages. Edited by Merv on 17 April 2011 at 9:45pm 1 person has voted this message useful | ||||||||||||
| tractor Tetraglot Senior Member Norway Joined 5251 days ago 1349 posts - 2292 votes Speaks: Norwegian*, English, Spanish, Catalan Studies: French, German, Latin | Message 5 of 10 17 April 2011 at 10:04pm | IP Logged | |||||||||||
No, from the article in The Economist.
That's my suspicion too. Edited by tractor on 17 April 2011 at 10:06pm 1 person has voted this message useful | ||||||||||||
| Doitsujin Diglot Senior Member Germany Joined 5118 days ago 1256 posts - 2363 votes Speaks: German*, English | Message 6 of 10 17 April 2011 at 10:50pm | IP Logged | |||||||||||
|
I find the whole idea interesting, but as anybody who ever took a statistics class can confirm, given a large amount of data you can prove pretty much anything if you pick the right correlations.
Here's one guy who apparently doesn't agree with Atkinson's math: Phonemic diversity decays "out of Africa"? Edited by Doitsujin on 17 April 2011 at 10:51pm 3 persons have voted this message useful | ||||||||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 5809 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 7 of 10 17 April 2011 at 10:55pm | IP Logged | |||||||||||
This forum is not hostile to religion. This forum is a discussion forum and is therefore a place for debate. There is no point in discussing something with someone who is not open to debate. Several of us here are genuinely interested in discussion the relative merits of language having a single root vs emerging simultaneously in different parts of the world. However, we cannot discuss that with you because you are not willing to follow any of the arguments that may be put forward. The prerequisites for debate are missing, so we can't have a debate. This is huuuuuuuuugely frustrating, because I would really love to discuss this topic. 5 persons have voted this message useful | ||||||||||||
| Merv Bilingual Diglot Senior Member United States Joined 5071 days ago 414 posts - 749 votes Speaks: English*, Serbo-Croatian* Studies: Spanish, French | Message 8 of 10 17 April 2011 at 11:51pm | IP Logged | |||||||||||
What are you even talking about? I follow "your" arguments, I just don't agree with them. Please, just stick to the two papers that Juan posted and stop addressing me in this tangential manner. If you want a debate, we can have one. I have explained why I thought Atkinson's paper was weak It's driven by ideology (much as one by me would be) that the origin of language must be Africa. It does not address any other parameters of how we might measure language evolution, namely vocabulary and grammatical structure. It just randomly jumps on phonemes, despite the fact that phonemic inventory can be a result of innovation just as much as a result of decay. Constructive grammatical innovation is much, much rarer than the innovation of new phonemes or even tonality. In fact, I'd be surprised to see any evidence of grammatical innovation that was not a compensatory change for an equal or greater loss in complexity from the parent language. Bulgarian, French, Hindi, Farsi, modern Greek, Norwegian all have demonstrated a radical loss in grammatical complexity from the ancestral languages. I would wager that that is the case with other modern languages, but would be interested in counterexamples. Edited by Merv on 18 April 2011 at 12:06am 3 persons have voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
